All the latest news from the industry weekly compiled by the editorial team for you free of charge.
This eMail is already registered.
An unexpected error occured.
Please accept our Terms of Use.
Registration successful.
1 Rating

HYPER-AUTOMATED MANUFACTURING Hyper-automated manufacturing - The new-age rope walk for manufacturing

Jan 18, 2022

With the change in global supply chains, India has a big door open to monetise from, especially given the US-China trade war. However, what holds us back in the world market is our hyper-reliance on manual labour, which is expensive, time-consuming and tedious. The only solution to overcome this roadblock is to make the right kind of investments into building hyper-automated factories. The Cover Story elaborates on why hyper-automated factories is a need of the time, the kind of changes it can bring into one’s shop floor & how it can open up more avenues for Indian manufacturing sector on the global platform. - Nikhil Ramaswamy, Co-founder & CEO, CynLr

The new year dawns on an optimistic note for the Indian manufacturing sector. Sure, there is still some uncertainty regarding when we can return to our pre-pandemic lives; however, most manufacturers are looking to capitalise on global seismic shifts in the supply chain. As global leaders look to diversify their manufacturing hubs, the time is ripe for India to capture and retain this interest. The only way to succeed is to grow aggressively while retaining the quality standards. But rising labour costs, supply chain interruptions, fierce global competition and natural disasters continually threaten to derail operations. More automation may be the answer, but the holy grail of manufacturing, ‘hyper-automated factories’, remain a pipe dream. How can manufacturers navigate this tightrope of scaling rapidly while future-proofing operations?

The key to successful tightrope walking, it turns out, is in building just the right amount of flexibility into strategic parts of the entire process. To achieve a good balance, tightrope walkers need to lower their centre of gravity. To achieve this, amateurs tend to lean forward on the tightrope, which puts them off balance. Professionals, however, stand straight but bend their knees slightly. This ‘give’ provides them stability and balance.

Similarly, a majority of the manufacturing industry still leans heavily on manual labour. Almost $15 trillion is spent worldwide on manual labour wages for low-skill, repeatable work (McKinsey report). These tasks are, ideally, prime candidates for automation. In contrast, the amount spent on automation is only $150 billion today. After the events of the past two years, C-suite executives are looking for balance.

Crossing the one big hurdle to be the next global manufacturing hub

A PwC study found that over 50% of CFO’s surveyed expressed an interest in accelerating automation. The increased intent to automate is reflected in the growth projections for industrial robotics. An October 2021 MarketStudy report found that the global industrial robotics market was worth $38.02 billion in 2020 and is expected to grow to $87.79 billion by 2028 at a CAGR of 10.35%.

The opportunity is now. There has been a decline in manufacturing goods from China to the United States due to political tension and trade wars. As a result, many global corporations are actively looking to future-proof their supply chains by diversifying offshore manufacturing locations. This is a massive opportunity for us, but the global manufacturing landscape has changed. India is not the only strong contender – even before the pandemic. Countries like Vietnam, Thailand, South Korea and Indonesia are hot on our heels. Indian manufacturing got only 10% of the $31 billion that moved from China to other Asian countries in 2019. 46% went to Vietnam and another 27% to Indonesia.

India may aim to be the preferred alternative manufacturing hub to China, but we have some critical challenges to overcome. Our robot density is much lower than the global average of 126 robots per 10,000 employees. India’s robot density is only four robots per 10,000 employees. In contrast, China has a robot density higher than the global average at 187 units per 10,000 employees (IFR). More reliance on manual labour is not the answer. Manufacturing labour productivity in China is also almost four to five times higher than in India because of increased automation (McKinsey). To win, we cannot play catch-up. What we need is strategic investments in the right technology to jump lightyears ahead of the competition.

Another issue for Indian manufacturers to navigate is that some industries are starting to express a preference for reshoring to the US. This intent is boosted by the fact that the US government recently approved an infrastructure bill allocating $300 billion to ramp up local manufacturing. $50 billion was for the semiconductor industry, which was severely hit by supply chain disruptions in the pandemic.

In short, we have the opportunity and the will to be the next global manufacturing hub but need to offer scale, quality and reliability. Ideally, we can ramp up and meet global levels of robot density, which also helps us improve labour productivity. However, the business case for robotics does not work for many Indian manufacturers today. Rising costs, high turnover, and other labour-related issues mean merely adding more people is not the solution.

Lack of adaptability: The real hurdle hindering India to improve robot density

In 1974, Philippe Petit successfully walked a tightrope between the two towers of the World Trade Center in the United States. The death-defying skywalk took intense behind-the-scenes preparation, which included training his own muscles to be fit, finding the right wire for the job, and connecting it between the towers with the right amount of tension. In an ideal scenario, the wire should have been extremely tight to eliminate any movement under the feet. However, to make his cable work for the real world – where rope construction, winds, human error and many other factors come into play – Petit built a little bit of slack for adaptability.

Similarly, most manufacturers know that highly rigid systems are bound to fail in less-than-perfect situations, which happens most of the time. Product designs are revised and changed periodically. If a factory cannot handle even minor changes, it becomes less desirable in the eyes of a global customer. Apart from minor changes to the design, product lifecycles are also getting shorter and shorter. With the technology available today, manufacturers need product life cycles of almost seven years to make the ROI work for automation, an unrealistic expectation nowadays.

Unfortunately, the industrial robots of today are too rigid and can perform only one task on a specific part placed in one particular orientation. The initial set-up involves heavy customisation. All further deviations or even minor changes in a product or product line also need more expensive customisation. For many, this type of disruption is not worth the time, money, effort, resources and stress.

So, manufacturers have resigned themselves to relying on manual labour to perform simple tasks like picking a screw from a bin, orienting it accurately and placing it in a specific way for a robot to perform its task. As a result, most robots today are concentrated in the distribution and e-commerce space to ‘transport’ shelves and deliver them to packing stations for speed sorting of parcels or delivery of totes and raw material. The bulk of tasks in manufacturing and individual piece-picking in warehouses are still manual because robots cannot intelligently see and grasp objects like human beings.

Humans are still required to perform very low-skilled, repeatable tasks like picking up objects from a bin and orienting and placing them correctly. These activities, known as ‘pick and place’ tasks, are too easy for humans but too complex for robots. The feasibility to automate these tasks drops to less than 50% when there is a hint of variability or uncertainty involved.

The productivity loss is enormous when we consider the sheer amount of pick and place tasks that most common products need. As an example, let us look at the use of fasteners - screws, bolts, nuts, washers and anything else used to ‘fasten’ one part to another. An average car has 3,500 fasteners. In 2021, 70 million automobiles were produced globally. Even if we estimate five seconds per fastener for a human to do the job, almost 340 million man-hours of effort are required annually to perform this simple object manipulation task for 70 million cars. In aerospace, a Boeing 747 aircraft uses approximately three million fasteners, almost 850 times that of a car. This is almost 4,166 man-hours per Boeing 747 aircraft to just pick, place and orient these fasteners. This high estimate is also an impractical scenario that expects every employee in a factory to perform their task in five seconds without any deviation, disruption, or error. Humans cannot sustain the same speed, accuracy, quality and consistency for low-skilled, repetitive tasks as robots can.

Another example where manual labour is employed for very simple pick-and-place tasks is that of machine tending for CNC machines. A human being is needed to pick a product out of a bin and place it the right way on a CNC machine. Even if a manufacturer were to invest in expensive customisation for their machine tending needs, a human is needed to prepare a structured pallet for the next robot in the assembly line.

So, the industrial robotics of today are too rigid and manual labour is too unreliable for low-skilled, repetitive tasks. What can a manufacturer do to stay competitive and scale quickly and efficiently?

We need robots that can see and understand objects the way humans do. We may have to take a page of the daredevil tightrope walker and invest in robots that have adaptability and flexibility built-in by design. Only then will manufacturers achieve a high ROI from their investments in industrial robots.

Robots that can see and think like humans can improve ROI

In an ideal world, 70-80% of the upfront investment a manufacturer makes on a robot should be reusable. Today, the actual robot arm is only 20-30% of the upfront costs, and 70-80% of the upfront costs are customisation costs. Costs like additional equipment, re-calibration, end-of-arm tooling, skilled specialists and testing can cost up to five times the cost of the robotic arm. These customisations are also very task and part specific. Minor changes often require the manufacturer to reincur customisation costs. The robot arm should be readily adaptable to new products, design changes and minor changes in orientation.

A robot arm needs to identify the right object in a bin, pick it up, reorient and place it in the right location. This type of intelligent flexibility will help manufacturers use the arm over a ten-year life cycle and realise a better ROI. The flexibility will give them the peace of mind to adapt their investment for new products, new customers and any other changes in design or process. This consistency will, in turn, allow Indian manufacturing to scale with an eye for quality control.

Universal automation is within an arm’s length, if we only reach for it

There is good news for manufacturers: advances in the field of machine vision may be the answer to inflexible robot arms, low ROI from industrial automation and manual labour challenges for low-skilled and repetitive tasks.

Robots using advanced machine vision can significantly leapfrog over current technology. These robots will be able to visualise, grasp and manipulate objects like humans. These intelligent robots will also introduce just the right amount of flexibility and adaptability to manage changes in design and process. In the words of Petit, “You must not force yourself to stay steady. You must move forward.” Robots that can see, understand and operate like humans will propel the world of manufacturing forward and help us realise the dream of hyper-automated factories.

Image Gallery

  • Manufacturing industry still leans heavily on manual labour. Almost $15 trillion is spent worldwide on manual labour wages for low-skill, repeatable work.

  • Most robots today are concentrated in the distribution and e-commerce space to ‘transport’ shelves and deliver them to packing stations for speed sorting of parcels or delivery of totes and raw material

  • A robot arm needs to identify the right object in a bin, pick it up, reorient and place it in the right location. This type of intelligent flexibility will help manufacturers use the arm over a ten-year life cycle and realise a better ROI.

  • Nikhil Ramaswamy

    Co-founder & CEO

    CynLr

Companies related to this article
Related articles